Molecular characterization of porcine epidemic diarrhoea virus (PEDV) in Poland reveals the presence of swine enteric coronavirus (SeCoV) sequence in S gene

Porcine epidemic diarrhoea (PED) is a highly contagious enteric viral disease of pigs with a high morbidity and mortality rate, which ultimately results in huge economic losses in the pig production sector. The etiological agent of this disease is the porcine epidemic diarrhoea virus (PEDV) which is an enveloped, positive single-stranded RNA virus. The aim of this study was to perform molecular characterization of PEDV to identify the strains circulating in Poland. In this study, 662 faecal samples from 2015 to 2021 were tested with reverse transcription quantitative real-time PCR (RT-qPCR) and the results showed that 3.8% of the tested samples revealed a positive result for PEDV. A phylogenetic analysis of the complete genome and complete S gene sequences showed that Polish PEDV strains belonged to the G1b (S-INDEL) subgroup and were closely related to the European PEDV strains isolated from 2014 to 2019. Furthermore, RDP4 analysis revealed that the Polish PEDV strains harboured a recombinant fragment of ~400 nt in the 5’ end of S gene with PEDV and swine enteric coronavirus (SeCoV) being the major and minor parents, respectively. Antigenic analysis showed that the aa sequences of neutralizing epitopes were conserved among the Polish PEDV strains. Only one strain, #0100/5P, had a unique substitution in the COE epitope. However, Polish PEDV strains showed several substitutions, especially in the COE antigen, as compared to the classical strain CV777. To the best of our knowledge, this is the first report concerning the molecular characterization of porcine epidemic diarrhoea virus strains, as well as the first phylogenetic analysis for PEDV in Poland.


Introduction
Porcine epidemic diarrhoea (PED), an acute and highly contagious enteric disease of pigs is characterized by watery diarrhoea, vomiting, dehydration and weight loss. The disease affects pigs of all ages but the most susceptible are neonatal piglets among which mortality can reach 100%, resulting in substantial economic losses [1][2][3]. The etiological agent of this disease is veterinary supervision, therefore no ethics committee approval was required [22]. Verbal informed consent has been obtained from all owners prior to the collection of faecal samples from the pigs.

RNA extraction and PEDV identification
The total of RNA was extracted from 140 μl of the collected supernatant using a QIAMP Viral RNA Mini Kit (Qiagen, Germany) according to the manufacturer's protocol. Extracted viral RNA was subjected to triplex RT-qPCR using a VetMAXTM PEDV/TGEV/SDCoV Kit (Thermo Fisher Scientific, USA) according to the manufacturer's recommendations. For each separate reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) run two positive controls (one for the RT-qPCR components and another for the RNA purification process) as well as one negative control (nuclease-free water, Ameresco, USA) were included. The RT-qPCR were run on the Mx3005P qPCR System and Aria MX (Agilent Technologies, USA) according to the manufacturer's instructions. A positive control was used to set the cycle threshold (Ct) for evaluating the test results. A VetMAXTM PEDV/TGEV/ SDCoV Kit was used to detect PEDV infection and exclude a potential PEDV-transmissible gastroenteritis virus (TGEV) and PEDV-porcine deltacoronavirus (PDCoV) co-infection [23].

NGS and Sanger sequencing
Total RNA extraction of the PEDV positive samples were subjected to Next Generation Sequencing (NGS) using Illumina's sequencing-by-synthesis (SBS) technology. For samples which could not be sequenced by NGS, sequences of the spike gene (S gene) were obtained by sequencing in both directions using the Sanger methodology. Four overlapping fragments were amplified with the primers described in Table 1 and the OneStep RT-PCR Kit (Qiagen, Germany). The reaction was conducted under the following conditions: 50˚C for 30 min, 95˚C for 15 min, 40 cycles at 95˚C for 40 s, 55-58˚C for 1 min and 72˚C for 1 min, followed by a final extension step at 72˚C for 10 min. The reverse transcription polymerase chain reaction (RT-PCR) products were visualized under ultraviolet (UV) light after electrophoresis in a 1.5% agarose gel, containing SimplySafe (EURx, Poland) in 1 x Tris-acetate-EDTA (TAE) Buffer (40 mM Tris/acetate buffer and 1 mM ethylenediamine tetraacetic acid (EDTA), pH 8.0). The RT-PCR products were purified using NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, Germany) and sequenced on a 3730xl DNA Analyser (Applied Biosystems, USA) using a Big Dye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, USA). The NGS and Sanger sequencing was performed by a commercial company (Genomed S.A., Warsaw, Poland). All of the novel sequences reported in this study were submitted to the Gen-Bank database under  . The beginning and end breakpoints of the potential recombinant sequences were also defined by the RDP4 software. Putative recombinant events were considered to be significant when P 0.01 was observed for the same event using four or more algorithms.

Phylogenetic analysis
The complete genome sequences of the 7 Polish strains obtained in this study were compared to other representative PEDV and swine enteric coronavirus (SeCoV) strains deposited in GenBank. The phylogenetic tree (Fig 1) showed that all Polish PEDV sequences were allocated, with a bootstrap value of 100%, within the G1b genogroup, together with sequences of European, Asian and American PEDV S-INDEL strains. They clustered in a branch which is clearly The nucleotide identity varied between the full length S gene sequences of 12 Polish PEDV strains within a range of 98.4% to 100% while the sequence identities between the Polish PEDV isolates and representative PEDV and SeCoV strains varied from 94.0% to 100% and from 91.1% to 94.2%, respectively. The phylogenetic tree (Fig 2) based on the full length S gene sequences confirmed that Polish PEDV strains were allocated within G1b subgroup. This affiliation was supported by a high bootstrap value (94%). Three clusters within the G1b subgroup were identified from Polish PEDV strains on the basis of the S gene sequences. The first was formed by sequences of three Polish isolates (#6706/1, #6220 and #6706/2) which originated from 2016 together with sequences of European, Asian and American PEDV S-INDEL isolates. These Polish PEDV isolates showed the closest sequence identity with the PEDV sequence from Austria (LT900502), France (KR011756), Spain (MN692763), Germany (LT898420), Slovenia (KU297956) and the Netherlands (MF974246) (99.7%-99.9%). The second cluster included two Polish strains (#25364/1 and #25364/2) from 2015 which originated from one herd which showed the closest sequence identity with the sequence of the Hungarian strain #HUN/5031/ 2016 (KX289955) and Slovenian strain #SLOreBAS-1/2015 (KY019623) (99.9%-100%). Finally, the third cluster was formed by 7 sequences of Polish PEDV strains (#44176/2, #44176/1, #0100/ 3T, #0100/4T, #0100/2M, #0100/5P and #0100/1L) located together with the Hungarian, French and Spanish strains (99.3%-99.8% nucleotide identity). Furthermore, an additional analysis of the partial sequences of the S gene revealed that the Polish strains #0100/3P, #0100/4P, #0100/ 4M, #0100/3L and #0100/5T belonged to the third cluster within the G1b subgroup together with strains #44176/2, #44176/1, #0100/3T, #0100/4T, #0100/2M, #0100/5P and #0100/1L (S1-S4 Figs). It is worth mentioning that all of the isolates identified in this third cluster correspond to the isolates obtained from the same herd between 2016 and 2017.

Deduced amino acid sequence analysis of neutralizing epitopes in the S protein
In order to study the genetic characteristics of the Polish PEDV strains, the deduced amino acid (aa) sequences of the S protein were compared with other reference PEDV strains representing the G2 genogroup (Non-S-INDEL strains) and the G1genogroup (S-INDEL-G1b and classical strains-G1a). We identified all 4 major epitopes capable of inducing neutralizing antibodies, the COE region, YSNIGVCK (SS2), SQSGQVKI (SS6) and GPRLQPY (2C10). In our study, sequences of epitopes SS2, SS6 and 2C10 were well conserved. As shown in Fig 3, compared to classical strain CV777, the Polish PEDV strains had the Y/S substitution at the third position in the SS6 epitope and the R/V substitution at the third position in the 2C10 epitope. Epitope SS2 was almost identical in all analysed strains. In the COE region (Fig 4)

Discussion
PEDV has become an important diarrhoea virus, causing serious economic losses for the pig industry worldwide. In Poland, the clinical symptoms of PED and the presence of specific antibodies against PEDV have been observed in several farms in 2015-2017 but no studies were performed to confirm the presence of PEDV. This paper, for the first time, presents data concerning the molecular characterization of PEDV strains from Poland.
Based on a phylogenetic analysis, PEDV is mainly divided into two genogroups, G1 (including the G1a and G1b subgroups) and G2 (with the G2a and G2b subgroups) [41,42]. In this study, a phylogenetic analysis was performed on the basis of whole genome sequences and showed that the Polish PEDV strains belonged to subgroup G1b and closely resembled the European PEDV S-INDEL strains isolated from 2014 to 2019. They were within the range of 99.6%-100% in terms of nucleotide identity. The phylogenetic tree based on the full length S gene, including more sequences of the Polish strains, confirmed the occurrence of the only G1b strain suggesting that this strain may be dominant in Poland. Furthermore, phylogenetic analysis allows for the classification of the Polish PEDV strains into three clusters with some geographical relationships. The first cluster included three strains isolated from 2016 and originated from two distinct farms located in neighbouring voivodeships (Dolnośląskie and Wielkopolskie). The second and third cluster comprised the strains from two herds-also located in neighbouring voivodeships. The second cluster contained two strains from 2015 which originated from one herd located in the Wielkopolskie voivodeship while the isolates identified in this third cluster originated from one herd from Kujawsko-Pomorskie voivodeship. It remains  [45], Italy [46,47] Austria [48], Portugal [49], Slovenia [50], and Hungary [51], which may suggest that it spread via a transboundary route. In this period of time, an increase in swine imports, as well as pork and pork products to Poland from western Europe (mainly: Germany, Belgium and Denmark) was noted [52,53]. This situation was correlated with the first detection of African swine fever (ASF) in Poland in 2014 [54] which adversely influenced the recovery of swine production in the following years. The subsequent increased importation of breeding stock from Denmark and Germany to help swine inventories recover and improve genetics [53] could also promote PED introduction to Poland. PEDV is mainly transmitted via the faecal-oral route, but shedding in semen has also been documented [55] and could therefore be a possible source of strains from western Europe, identified in this study. Moreover, indirect transmission via contaminated fomites, as well as cross-contamination via feed cannot be excluded, as was confirmed previously [56,57].
It has been shown that PEDV S-INDEL strains belonging to the G1b subgroup have a relatively low degree of pathogenicity and cause mild symptoms of the disease compared to the

PLOS ONE
highly virulent non-INDEL strains [1,17,58]. For this reason, infection with these strains may be overlooked which additionally complicates the accurate estimation of PED prevalence in Poland. During the epidemic in Asia and America the presence of both genotypes have been confirmed, while in Europe the presence of the non-INDEL variant was only identified in the Ukraine [59]. In other European countries only the occurrence of a variant with a relatively low degree of pathogenicity (S-INDEL) was recorded [17,24,55]. However, because PED is not a significant disease in the EU and it is not amongst the World Organisation for Animal Health (OIE) listed diseases, to date most countries have not implemented active monitoring for this particular disease, so the information concerning the currently circulating PEDV variants is limited or unknown, facilitating the subsequent spread of the virus. Moreover, the appearance of further genetic variations should be expected as PEDV strains are especially prone to mutations and recombinations [4, 60,61].
The S gene is commonly used as the target gene in studies concerning the genomic characteristics of PEDV strains. The PEDV S protein is responsible for receptor binding and viral entry, and thus, determines the host range and cell tropism [62,63]. In addition, neutralizing epitopes have been found in the S protein thus this protein is a primary target for vaccination against PEDV [4,64]. Neutralizing antibodies play an important role in the prevention and control of viral infection therefore it is important to analyse changes in their amino acid sequences. Four neutralizing epitopes in the PEDV S protein have been determined, COE [65], 2C10 [66], SS2 and SS6 [67], and all 4 were detected in sequences of the S gene of the Polish PEDV strains. An antigenic analysis showed that the aa sequences of all epitopes were conserved among the Polish PEDV strains. Only one strain, #0100/5P, had a unique substitution in the COE epitope. However, the Polish PEDV strains showed several substitutions, especially in the COE antigen, as compared to the classical strain CV777. The same differences in residues were observed in other S-INDEL strains from the G1b subgroup as well as in most of the non-INDEL strains belonging to genogroup G2. Mutations in the sequences of epitopes may alter the antigenicity, pathogenicity and neutralization properties of strains [68]. Therefore, a vaccine derived from the prototype strain CV777 protects against the disease caused by classical strains [69] but does not provide adequate immune protection against the currently emerging strains [58,[70][71][72]. Recent vaccine strains may only partially induce neutralizing antibodies against emergent PEDV strains which pose a major challenge to the prevention and control of PED.
Recombination plays a pivotal role in the diversity and evolution of coronaviruses by creating new strains with altered virulence [73]. Several reports have identified recombinant sequences in the PEDV S gene, ORF1a and ORF1b [69,74]. Since 9 of the Polish PEDV strains are clustered together with recombinant PEDV-SeCoV isolates described in Hungary, Italy, Spain and Slovenia [24, 51, 75, 76] we verified whether or not Polish PEDV strains also resulted from such a recombination event. RDP4 analysis confirmed that Polish PEDV strains harbour a recombinant fragment of~400 nt in the 5' end of the S gene with PEDV and SeCoV being the major and minor parents, respectively. Putative recombination events were detected using all 7 statistical methods with a high degree of significance and reliability. Our results indicated that recombination is a common phenomenon among Polish field PEDV strains since 9 out of 12 of the analysed strains represented a recombinant PEDV-SeCoV variant. The clear separation of the PEDV-SeCoV isolates from the other European strains, suggests a new independent evolution of PEDV in Europe from 2015 due to a recombination event in the Sgene between PEDV and SeCoV. Detection of new chimeric (recombinant) coronavirus affecting pigs called swine enteric coronavirus (SeCoV) should clarify the origin of the novel recombinant PEDV isolates. Most of the genome of this new chimeric virus is derived from TGEV, but the S-gene is derived from PEDV [75]. SeCoV was reported in several European countries including Italy [75], Germany [77], Spain [de Nova] and countries from Central Eastern Europe [78,79]. It causes the same clinical signs as PEDV and TGEV and because their recombinant nature, a diagnosis based on the detection of a particular sequence for both PEDV and TGEV may lead to misidentification and the presence of SeCoV may be unnoticed [de Nova]. In Poland, no studies have been performed to confirm the presence of SeCoV.
In conclusion, the Polish PEDV strains identified in this study clustered into a G1b subgroup and were closely related to the European PEDV S-INDEL strains. Compared with the prototype strain CV777, Polish PEDV strains had multiple variation in neutralizing epitopes, suggesting that the development of a novel vaccine may be necessary for the control of PED in Poland. Furthermore, in this study, a natural recombination event involving a~400 nt fragment of SeCoV was identified in the Polish PEDV strain. To the best of authors' knowledge, this is the first report concerning the genetic characteristics of the virus in Poland. These results provide valuable information concerning PEDV strains circulating in the country which is especially important for the effective control of the disease and limiting the losses in swine production.